Recent studies indicate a decline in biopharma clinical trial success rates over the past several years. An analysis conducted by the Biotechnology Industry Organization (BIO) and BioMedTracker showed that from 2003 to 2010, only 9 percent of drugs made it from phase I to Food and Drug Administration approval, versus about 17 percent to 20 percent in previous years.
Investigators also have identified variations in trial success rates that depend on therapeutic area and molecule type (i.e., small molecule or biotherapeutic). These findings may help elucidate the contributing factors that underlie changes in clinical trial attrition rates.
Generally, proposed etiologies concern issues with disease complexity, clinical trial designs, therapeutic area endpoints, regulatory scrutiny, and globalization. Recommended remedies primarily relate to alternative trial models, such as various proof-of-concept studies, translational medicine, and adaptive trials.
Analyzing the bases for clinical trial attrition rates is an inexact science, to say the least, as there are multiple confounders and long timelines. It is thus difficult to pinpoint causes with any certainty. Nonetheless, the stakes could not be any higher, given the well-recognized financial challenges facing biopharmas. Furthermore, many believe that clinical trial approaches have not kept current with other aspects of the industry’s evolution and maturation, so the time is right for developers to initiate major reforms.
The Problem—Clinical Trials Failing More Often
According to Tufts Center for the Study of Drug Development (Tufts CSDD), the average cost for developing a new drug is now about $1.3 billion, and the average time required to conduct trials and obtain regulatory approval is seven years.
Reflecting on the gravity of the situation, Tufts CSDD Director Kenneth I. Kaitin said, “The research-based drug industry, in the United States and globally, is not sitting still, but the question remains whether developers can bring enough new drugs to market at the pace needed to remain financially viable.”
The February 2011 BIO/BMT report described lower-than-expected overall clinical trial success rates. The analysis was quite comprehensive, covering 4,275 drugs and 7,300 indications for all types of therapeutics in all phases of FDA clinical trials from late 2003 to the end of 2010. These drugs were being developed by all kinds of privately and publicly owned pharmaceutical and biotechnology companies.
Key findings from the BIO/BMT study included:
- The success rate of bringing new medicines to market in recent years is only about half of what it had been previously (from 2003 to 2010, 9 percent of drugs made it from phase I to FDA approval, versus about 17 percent to 20 percent in previous years).
5 J.A. DiMasi, L. Feldman, A. Seckler, and A. Wilson, “Trends in Risks Associated With New Drug Development: Success Rates for Investigational Drugs,” Clinical Pharmacology & Therapeutics, Vol. 87, No. 3, March 2010.
- Biologics have a much higher probability of reaching approval than do small molecules—biologics’ overall success rate was 26 percent for lead indications, versus an overall rate of 14 percent for new molecular entities’ lead indications (NMEs, or small molecule drugs).
6 Id.
- Success rates by phase:
7 Tim Casey, “Drug Approvals Fall in 2010,” First Report Managed Care, March 18, 2011.
- ▸ 15 percent make it from phase II trials to approval,
- ▸ 44 percent make it from phase III trials to approval,
- ▸ 80 percent submitted for approval are approved.
- Success rates at phase III by therapeutic area:
8 Michael Hay et al., “BIO/BioMedTracker Clinical Trial Success Rates Study,” presentation at BIO CEO & Investor Conference, Feb. 15, 2011.
- ▸ the autoimmune area was the most successful at 63 percent;
- ▸ the least successful area was oncology, with only 34 percent reaching approval;
- ▸ the respiratory area had a 61 percent success rate;
- ▸ the endocrine area had a 60 percent success rate;
- ▸ the infectious area had a 55 percent success rate;
- ▸ the neurology area had a 55 percent success rate;
- ▸ the cardiovascular area had a 46 percent success rate.
Michael Hay, senior biotechnology analyst at BMT, commented, “Strikingly, oncology drugs have the toughest time making their way through the clinic, despite cancer being the most closely studied area in drug development.”
Another study by investigators from Tufts CSDD looked at a different population of trials and drug developers, as well as dissimilar time periods, but reported several similar findings. The analysis examined drugs in the pipelines of the 50 largest pharmaceutical companies (by sales) that had their first entry into trials from 1993 through 2004.
These researchers also noted a decline in clinical development success rates. They reported in 2010 that “the overall estimated clinical approval success rate is lower than it has been for prior periods.”
- For all compounds the clinical approval success rate for the entire study period was 19 percent.
- Large molecule drugs (biologics) have twice the success rate of small molecule drugs.
- ο The chance of moving from phase I all the way to FDA approval was 32 percent for large molecules and only 13 percent for small molecules.
- “The top 10 pharmaceutical companies out of the world’s top 50 have lower estimated overall clinical approval success rates than do smaller firms in that group … . Despite experiencing lower overall clinical success rates, the top 10 firms terminated a greater proportion of their failures in early stage clinical testing, compared to the other 40 companies in the group, the study found.”
12 “Approval Success Rates Higher for Smaller Firms Among Top 50 Pharmaceutical Companies,” press release, Sept. 9, 2010, Tufts Center for the Study of Drug Development.
- “The estimated clinical approval success rates and phase transition probabilities differed significantly by therapeutic class.” There are a few differences between this study and the BIO/BMT study that make direct comparisons difficult. For example, this study did not break out the oncology area separately, but combined it with immunology. Also, the Tufts CSDD analysis by therapeutic area only looked at drugs that originated within the pharmaceutical developer, eliminating in-licensed drugs. The results for success rates passing from phase III to regulatory review are:
- ▸ respiratory was the most successful area at 85.7 percent:
- ▸ central nervous system (CNS) was the least successful area at 46.4 percent;
- ▸ the musculoskeletal area had an 80 percent success rate;
- ▸ the systemic anti-infective area had a 78.6 percent success rate;
- ▸ the cardiovascular area had a 64.3 percent success rate;
- ▸ the antineoplastic/immunologic area had a 55.3 percent success rate.
Finally, CMR International’s “2010 Pharmaceutical R&D Factbook” found that “the number of experimental drug projects terminated at the final Phase III stage of development had doubled in the period 2007-2009 compared with 2004-2006.”
The Suspected Etiologies
Unraveling what causative factors might be driving these disappointing clinical trial outcomes is difficult given the multidimensional nature of these analyses. There certainly are a number of issues at play, and they likely vary by setting and in importance. Sources of potential etiologies include:
- disease complexity,
- study designs,
- endpoints,
- regulatory oversight, and
- globalization.
Now that the pharmaceutical industry has matured, therapies for most of the “low hanging fruit” diseases have been developed and marketed, leaving only complex, poorly understood conditions to be tackled. Allan Haberman, founder of the Biopharmaceutical Consortium, wrote, “A large part of the reason for developmental attrition is that companies have been increasingly addressing complex diseases with high unmet medical needs. Addressing these complex diseases involves addressing unprecedented targets. However, drugs that address unprecedented targets are much more likely to fail in Phase II (by a factor of 2- to 4-fold) than drugs that address precedented targets.”
Charles Gombar, Ph.D., vice president of R&D strategy and business improvement at Wyeth, commented, “On the Phase II attrition issue, I think that what we saw over the past decade was a big change in the drug development game. Part of that change stemmed from the fact that you must have novel drugs. You have to bring new value and real medical value to the marketplace. That naturally drives people to novel targets, which carry higher risk and much more uncertainty. I’m not sure whether Phase II attrition is really an issue or simply a manifestation of how the drug development game has changed.”
There are a number of study design problems that might drive high trial failure rates, including preclinical studies that lack robustness, suboptimal evaluations of efficacy, and clinical trial protocols with excessive complexity.
P.R. Lowenstein and M.G. Castro from the University of California at Los Angeles and Cedars-Sinai Medical Center explained how preclinical studies that are narrow in scope can be misleading: “We propose that a likely cause of such failures is the lack of ‘robustness’ in the preclinical science underpinning the Phase I/II and III clinical trials. … Many times preclinical experiments are tested in a very narrow set of experimental conditions. Thus, when such approaches are finally tested in the context of human disease, the challenge provided by the varied age of patients, the complex genetic makeup of human populations, and the complexities of the diseases to be treated provide challenges which were never tested or modeled.”
These investigators also pointed out that early clinical trials are designed primarily to evaluate toxicity and safety issues. Substantive tests of efficacy are left to Phase III trials, at which point major investments in time and money have been made.
Haberman’s work also identified preclinical studies as a problem. “Poorly predictive animal models constitute a major cause of drug attrition,” he reported. He also pointed to CNS and cancer as particularly demonstrative of “therapeutic areas in which animal models are notorious for being poorly predictive.”
Tufts CSDD has observed that protocols have increased in complexity in terms of the number of procedures required and the burden of execution. A study conducted by Tufts reported that “the median number of procedures per clinical trial increased by 49 percent between 2000-03 and 2004-07, while the total effort required to complete those procedures grew by 54%.”
A number of researchers and industry analysts point to problems with endpoints in specific therapeutic areas, especially oncology, as a driver of lower trial success rates. For example, John Craighead, Ph.D., of BIO suggested that oncology trial efficacy endpoints in phase II trials are insufficiently predictive of phase III success.
Anastassios D. Retzios, a clinical trial consultant with extensive pharmaceutical development experience, agreed: “A major contributor to the high failure rate is inadequate Phase 2 programs that provide sub-optimal information for the “go/no go” decision to move to Phase 3 and the design of the Phase 3 trials.”
A group of academic oncology investigators made the point that newer cancer drugs require different phase II primary endpoints than did traditional, cytotoxic agents. “Although traditional oncology trial designs using the endpoint of response and a single arm design seem to have done this task reasonably well for cytotoxic agents, the same does not seem to be true for newer agents in which high rates of tumor shrinkage may not be expected, nor for combinations of agents (such as a new drug combined with standard treatments). Certainly, success rates for phase III trials seem to be decreasing.”
FDA regulatory oversight is another area cited as a potential cause of higher attrition rates. Craighead of BIO said, “Many people would argue the safety hurdles to get a drug approved are much higher than they used to be. … The bars for approval have increased, especially in the area of safety.”
Finally, some propose that the industry’s evolution to global markets might complicate clinical trials. Contract research provider Jean-Pierre Tassignon asserted that pharmaceutical developers’ dislocation of clinical trials to overseas markets has lowered trial quality and may be contributing to failure rates. “In 2007-2009, North American and European sites have seen their respective shares in Phase III trials decline in favor of the rest of the world. … [D]ata from healthcare systems with health outcomes significantly worse than in North America and Europe comes cheaper, but may jeopardize the overall success rate of new drug projects coming out of Phase II.”
The Proposed Solutions—New Approaches to Trials
There are a multitude of prescriptives offered for improving clinical trial success rates, almost all of which pertain to modifying the types, sequencing, or design of studies conducted throughout the development process.
For example, Lowenstein and Castro propose that preclinical studies should put greater emphasis on efficacy and be expanded for robustness. They wrote, “Especially the efficacy ought to be challenged by testing the therapeutic approach in multiple animal models of varied genetic backgrounds.”
New study designs being pursued by biopharmas include exploratory Investigational New Drug applications (INDs), Phase 0 human clinical studies using microdosing, and adaptive trials. Exploratory INDs are pre-phase I FDA , and involve testing of up to five chemical entities or formulations simultaneously to identify a lead compound. Testing more candidates in humans promotes better selection of which drug to advance.
Microdosing studies involve testing humans with subpharmacological, trace doses of drug to obtain basic pharmacokinetic (PK) data. This human PK information “can then be used to (a) assist in the candidate selection process, (b) determine the first dose for the subsequent phase I study on the selected candidate, (c) establish the likely pharmacological dose and (d) calculate the likely cost of goods.”
Adaptive trials allow for changes to some aspect of the trial protocol based on the data gathered as the trial progresses. “Adaptive designs have the potential to improve drug development when appropriately applied. … In general, adaptive trial designs offer increases in information relative to cost or improved patient care within a trial,” wrote Gaydos et al.
Translational medicine, beginning with translational research, is advocated by most industry participants. “Translational medicine programs have been widely adopted in the pharmaceutical industry, and most drug development experts believe that these programs are the best approach to reducing attrition and thus to reducing costs and improving productivity of drug development.”
“Translational medicine is the integrated application of innovative pharmacology tools, biomarkers, clinical methods, clinical technologies and study designs to improve disease understanding, confidence in human drug targets and increase confidence in drug candidates, understand the therapeutic index in humans, enhance cost-effective decision making in exploratory development and increase phase II success,” wrote B.H. Littman et al.
One key tool in translational medicine is the proof-of-concept (POC) trials study, which involves using a biomarker instead of a clinical endpoint, and testing in a small number of patients to get an early estimate of efficacy. Biomarkers also have been widely recommended as a means for advancing clinical trials, and have been included in pharmaceutical development programs for a number of years now.
Yet pharmaceutical developers are still learning how best to develop and use biomarkers. Evan Loh, M.D., vice president of clinical R&D at Wyeth, stated, “Biomarkers and translational medicine are something the whole industry is struggling with, to be honest. With the biomarker concept, we are struggling to determine the true value that any biomarker gives us relative to increased confidence in decision-making.” Also, Peter Lassota, Ph.D., divisional vice president, imaging biology and oncology at Caliper Life Sciences, said, “We need better markers for early detection of disease, particularly cancers, in order to improve therapeutic outcomes. We are in need of surrogate markers of efficacy.”
A panel of biopharma executives and academics convened by Tufts CSDD in early 2011 recommended the early and carefully managed incorporation of biomarker development, as well as the use of emerging computing, telecommunication, and imaging technologies to support these translational science activities.
Finally, biopharmas seem to be responding to the complexity and logistical challenges of global clinical trials by returning to more of a domestic focus, according to Tufts CSDD. They report as part of their 2011 outlook on near-term trends, “Although more than half of all FDA-regulated clinical trials in 2010 were conducted outside the U.S., sponsors will seek to decrease the number of countries hosting development activity in an effort to reduce global logistical and regulatory complexity.”
Conclusions
One key distillation of clinical trial attrition problems leads to the need for better biologic models of the target diseases. With better models, and hence perhaps more accurate means of predicting efficacy as well as safety, the right drugs will be passed through to later phases, and there should be fewer late-stage surprises. Haberman wrote, “We conclude that the ability of researchers to successfully identify and validate biomarkers, and to design and carry out POC clinical trials, depends to a large extent on an understanding of disease biology and disease pathways. Thus, biology-driven strategies of drug discovery carry over to the new paradigm of early drug development.”
This theory is consistent with findings that biologic therapies, which typically have better-understood biologic models relative to small molecule drugs, outperform small molecules in clinical trial success rates.
Two trends may provide a sanguine outlook in terms of improved biologic disease models and thus lower trial attrition rates going forward. First, biopharmas’ recent launch into new types of open, closely knit R&D collaborations with academic medical centers might greatly advance translational medicine achievements.
Secondly, genomics and molecular diagnostics technologies are advancing at exponential rates, as are computing tools that enable powerful bioinformatics capabilities. These tools should greatly facilitate elucidation of the biology of complex disease, as well as provide new, more effective biomarker approaches.
In summary, it is likely that the recent observations regarding declining success rates compared to prior periods do not reflect the effectiveness of new initiatives such as translational medicine, biomarkers, adaptive trials, and other new study designs. Focusing and investing in deeper understanding of the biologic bases of the target disease, including developing models and tools that can be brought forward into trials, should enable earlier and more predictive efficacy testing and hence improved success rates.
Learn more about Bloomberg Law or Log In to keep reading:
Learn About Bloomberg Law
AI-powered legal analytics, workflow tools and premium legal & business news.
Already a subscriber?
Log in to keep reading or access research tools.